天津大学《大数据分析》课件04OLAP.pdfVIP

CS412Intro.toDataMining

Chapter4.DataWarehousingandOn-line

AnalyticalProcessing

JiaweiHan,ComputerScience,Univ.IllinoisatUrbana-Champaign,2017

1

2

Chapter4:DataWarehousingandOn-lineAnalytical

Processing

DataWarehouse:BasicConcepts

DataWarehouseModeling:DataCubeandOLAP

DataWarehouseDesignandUsage

DataWarehouseImplementation

Summary

3

WhatisaDataWarehouse?

Definedinmanydifferentways,butnotrigorously

Adecisionsupportdatabasethatismaintainedseparatelyfromthe

organization’soperationaldatabase

Supportinformationprocessingbyprovidingasolidplatformof

consolidated,historicaldataforanalysis

“Adatawarehouseisasubject-oriented,integrated,time-variant,and

nonvolatilecollectionofdatainsupportofmanagement’sdecision-making

process.”—W.H.Inmon

Datawarehousing:

Theprocessofconstructingandusingdatawarehouses

4

DataWarehouse—Subject-Oriented

Organizedaroundmajorsubjects,suchascustomer,product,sales

Focusingonthemodelingandanalysisofdatafordecisionmakers,notondaily

operationsortransactionprocessing

Provideasimpleandconciseviewaroundparticularsubjectissuesbyexcluding

datathatarenotusefulinthedecisionsupportprocess

5

DataWarehouse—Integrated

Constructedbyintegratingmultiple,heterogeneousdatasources

relationaldatabases,flatfiles,on-linetransactionrecords

Datacleaninganddataintegrationtechniquesareapplied.

Ensureconsistencyinnamingconven